INDEX

Explanations

questions or phrases regarding capability and action

oai_token-act-pair · gpt-4o-mini Triggered by @bot

can do

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 prof

-0.35

 foot

-0.32

 being

-0.32

 head

-0.31

 Hage

-0.31

the

-0.30

 layout

-0.30

 voet

-0.30

 regime

-0.29

日発売

-0.29

POSITIVE LOGITS

 canst

0.81

 JpaRepository

0.77

CanBe

0.75

 pinulongan

0.72

Can

0.68

IntoConstraints

0.67

 pouvons

0.66

Can

0.64

ניתן

0.64

 могу

0.62

Activations Density 0.018%

questions or phrases regarding capability and action

can do

No Comments

No Known Activations